Protein-protein interactions and NMR: G protein/effector complexes

Transcription

1 Protein-protein interactions and NMR: G protein/effector complexes Helen Mott 5th CCPN Annual Conference, August 2005 Department of Biochemistry University of Cambridge

2 Fast k on,off >> (ν free - ν bound ) A + B k on k off AB k on,off ~ (ν free - ν bound ) Slow k on,off << (ν free - ν bound ) να free να bound

3 15 N-HSQC: Titration of Unlabelled Peptide into Labelled Protein 1:0 1:0.1 Fast 1:0.3 Intermediate Slow? 1:0.5 1:1.0 Fast/ Intermediate

4 15 N-HSQC: Titration of Unlabelled Peptide into Labelled Protein 1:0 1:0.1 Fast 1:0.3 Intermediate Slow? 1:0.5 1:1.0 Fast/ Intermediate 1:1.2

5 Measure K d e.g. fluorescence, ITC, Biacore, NMR Minimize interacting region, if necessary e.g. limited proteolysis Find conditions where both components are stable and interaction still occurs (e.g. salt, ph dependence) and exchange regime is favourable Add unlabelled component in excess to saturate the other component Department of Biochemistry University of Cambridge

6 Information attainable depends on the strength of the complex Structures of protein complexes Dynamics of interactions Mapping interactions e.g. for mutagenesis studies Extracting distance information (transferred NOE) Docking strong strong weak or strong weak or strong weak or strong Department of Biochemistry University of Cambridge

7 Differential Labelling Schemes Labelled with 13 C, 15 N Unlabelled 12 C, 14 N 13 C H H 12 C Simplification of spectra 13 C H H 12 C Department of Biochemistry University of Cambridge

8 Cdc42 + Cdc42/ACK Complex ACK Free 15N-labelled ACK 15N-ACK + unlabelled Cdc

9 Tight Complexes: Slow Exchange (the ideal case) Chemical Shift Assignments 1. Backbone experiments: HSQC, HNCA, HN(CO)CA, HNCACB, CBCA(CO)NH recorded on each component of complex 2. Sidechain experiments: 15 N-separated TOCSY, HCCH-TOCSY CC(CO)NH, H(CCCO)NH, HBHA(CBCA)(CO)NH on each component of the complex 3. Coupling constant experiments: HNHA on each component of the complex

10 Tight Complexes: Slow Exchange (the non-ideal case) Chemical Shift Assignments 1. Backbone experiments require full deuteration and TROSY: HSQC, HNCA, HN(CO)CA, HNCACB, HN(CO)CACB, HNCO, HNCACO 2. Sidechain experiments: CC(CO)NH (on deuterated sample) On protonated sample: HCCH-TOCSY, 15 N-separated NOESY (to guess shifts of sidechain 1 H) 3. Coupling constants: experiments do not work, use all shifts (including CO) to put into Talos. Full set of assignments of the free protein(s) is necessary

11 NOE Data (in all cases) 2 types of experiment: Edited or separated: keep 1 H attached to nucleus X 13 C and 15 N-separated NOESYs NOEs between 13 C-attached protons and NOEs between 13 C-attached and 12 C- attached protons Filtered or rejected: keep 1 H NOT attached to nucleus X X-filtered NOESY NOEs between 13 C- 1 H and 12 C- 1 H, 14 N- 1 H Doubly-rejected NOESY/TOCSY 12 C- 1 H/ 14 N- 1 H

12 X-filtered Experiments Two basic types: a) Schemes where X-attached 1 H are removed directly 1 H 1/2J X e.g. generate MQC b) Schemes where interleaved experiments are recorded and have to be added or subtracted to keep or discard X-attached 1 H 1 H X X-pulse present or absent in alternating experiments

13 All 13 C-filtered experiments suffer because 1 J CH varies: Hz aliphatic CH; Hz aromatic CH 1 delays are inevitably only tuned to one J coupling J CH (Hz) aliphatic His Tyr, Phe, Trp δ 13 C (ppm) 1 J CH = (0.365 ± 0.01 Hz/ppm) δc ± 0.5 Hz 1 J CH = A δc + B Zwahlen et al (1997) JACS

14 Adiabatic pulses Trajectory of 13 C magnetization follows effective field for duration of the pulse Carrier frequency of the pulse starts far upfield, sweeps through resonance, then downfield 13 C resonate at different frequencies so they are all inverted at different times, depending on their frequency and the sweep rate and duration of the inversion pulse

15 Purge scheme with adiabatic pulse 1 H 13 C G a Hy + Iy G1 1 /2 1 /2 From a to b: t G1 b G2 13 C spin inverted at time t (t 0) after 1 H 180 pulse 1 J CH evolution occurs for a time 1-2t 1 J CH and t vary for each 13 C nucleus Hy + IycosπJ IS ( 1-2t) - 2IxSzsinπJ IS ( 1-2t) 90 ( 1 H) Hz + IzcosπJ IS ( 1-2t) - 2IxSzsinπJ IS ( 1-2t) IzcosπJ IS ( 1-2t) = 0 for all spins 1-2t = 1/(2 1 J CH ) (i) removed by G2 Since t 0 and 1 J CH is smallest for upfield 13 C (i.e. methyls), these must be inverted first. 1 is set close to 1/2 1 J CH (methyl) 1 J CH = Aδ C + B (ii) (A=0.365 Hz/ppm; B = Hz) Frequency of the transmitter δ RF (t) = δc. Substitute for 1 J CH in (i) 1-2t = 1/[2(Aδ RF + B)] time derivative to get sweep rate during WURST Zwahlen et al (1997) JACS

16 3D 13 C-filtered, 13 C-edited NOESY-HSQC ( X-filter ) Two of these modules at the start of the sequence Suppression factors are fold Putting filter modules at the start of the sequence makes H 2 O suppression easier No decoupling during T1 ( 1 H evolution) to detect breakthrough 15 N filter at the same time as the 13 C filter

17 2 Issues with the X-filter experiment 1. It has to be referenced in both proton dimensions 2. It has to be calibrated for structure calculation and there are no reference distances Department of Biochemistry University of Cambridge

18 Structure calculation strategies (1) Assignment of intra-molecular NOEs in 13 C- and 15 N-separated spectra: check that they do not have any inter-molecular possibility (they may not appear in X-filter if they are weak) Ambiguous NOEs that appear in 13 C-edited NOESY as well as X-filter, 13 C-edited NOESY should be treated in the same way as other ambiguous NOEs as they may contain intra-molecular contribution to the intensity Check for the other things that appear in the X-filter experiment: Thr/Ser-OH ~ 5ppm Tyr-OH ~9ppm Cys-SH ~1ppm! In larger proteins, X-filter only really works for methyl groups Watch out for the artefacts..

19 Structure calculation strategies (2) Dealing with all the data - generation of restraints minimum 5-6 NOESY datasets all with different contributing atoms: the possibilities must be edited for the different experiment types e.g. 13 C-edited NOESY: F1/F3 = 13 CH only; F2 = any 1 H Starting from two extended chains with ideal geometry not lying on top of each other (bias starting structures) Starting from the free structures do not bias initial intermolecular possibilities with orientation.

20 Cdc42-21 kda small G protein of the Rho family GTP cofactor replaced by GMPPNP PAK - 5kDa fragment - can be expressed as GST-fusion K d ~ 30nM Samples: 15 N Cdc42 + unlabelled PAK 15 N, 13 C Cdc42 + unlabelled PAK 15 N PAK + unlabelled Cdc42 15 N, 13 C PAK + unlabelled Cdc42 no deuteration required - all experiments worked 3D X-filtered, 13 C-edited NOESY ran on 13 C, 15 N PAK, unlabelled Cdc42

21 13 C-separated NOESY (intra- and intermolecular) 13 C-filtered, 13 C-separated NOESY (inter-molecular only)

22 Summary of Cdc42/PAK Intermolecular NOEs Unambiguous Ambiguous 13 C-filter/ 13 C-edited C-NOESY (PAK) C-NOESY (Cdc42) N-NOESY (PAK) N-NOESY (Cdc42) 7 12 Total Intermolecular Total number of NOEs = 4,000

23

24 HR1b domain of PRK1 complexed with Rac1 small GTPase HR1b Rac1

25 Hr1b/Rac1 (30 kda) Correlation time ~20 nsec 100% deuteration for backbone assignment of Rac (methyl protonation too expensive) HR1b 13 C/ 15 N labelling only Department of Biochemistry University of Cambridge

26 Methyl group intermolecular NOEs Department of Biochemistry University of Cambridge

27 X-filter artifacts -dispersive diagonal Department of Biochemistry University of Cambridge

28 0 0 Arg Cδ/Hδ 1 1 Hγ Hδ Hδ X-filter artifacts water exchange peaks - incomplete suppression of strong crosspeaks H 2 O Department of Biochemistry University of Cambridge

29 HR1b/Rac Very few NOEs in X-filter experiments on HR1b side (9 out of 62) HR1b proton shifts overlapped No aromatics in interface Only 2 Rac residues in the interface have methyl groups Interaction between two helices - no intermolecular NOEs in 15 N- separated NOESY ( 2 H Rac/ 1 H HR1b) Department of Biochemistry University of Cambridge

30 X-filter experiments work best for hydrophobic interfaces, with methyl groups and aromatics For larger complexes, -CH3 and flexible residues give NOEs (ILV approaches should help) For interfaces dominated by salt bridges the NOEs in the 13 C-separated X-filter experiment will be extremely sparse Intermolecular β-sheet interfaces - 15 N-separated NOESY on 100% 2 H/ 15 N labelled protein mixed with unlabelled partner Department of Biochemistry University of Cambridge

31 Cross-saturation Takahashi et al (2000) Nat. Struc. Biol Protein II Protein I Strong binding case R.F. -NH -CH - 15 NH -C 2 H Band-selective proton saturation, followed by TROSY-HSQC cross-saturation Saturate the aliphatics of protein II - magnetization transferred by spin diffusion to the aromatics and amides. Cross-saturation to the 15 NH in the interface on protein I Measure intensity in HSQC vs time of saturation to find residues in interface More precise than chemical shift mapping Drawbacks Cys-SH also saturated, transferred within protein I Spin diffusion between NH in protein I Department of Biochemistry University of Cambridge

32 Using NMR data for docking Structures of components are known Use NMR data to map binding contacts or determine relative orientation of components

33 Sec5 NMR Structure ~0.5Å RMSD C N

34 Sec5/Ral Titration - all slow exchange 1:0 1:0.5 1:1.0

35 Definition of Residues in the Contact Site Pick all the peaks in free and bound and calculate combined shift difference in 15 N and 1 H shifts, often defined as: [( 15 N) ( 1 H) 2 ] 1/2 Define significant chemical shift perturbation (>1 SD from average shift change) - add in the ones that have disappeared completely Check for solvent accessibility (e.g. NACCESS): NHs that are completely buried are unlikely to be involved in the interaction but are experiencing secondary effects, unless there is a significant structural change (should be obvious from the HSQC). NACCESS cutoff: if the residue is more than 50% exposed it is available for interaction.

36 Shift changes after accessibility filtering

37 Using Chemical Shift Mapping Data for Docking (HADDOCK) Take significant shift changes, screened by solvent accessibility Active residues Take residues close to active residues on the surface Passive residues Ambiguous interactive restraint (AIR) N atoms N res B N atoms ( ) (-1/6) d iab = Σ Σ Σ 1 d 6 m ia =1 k=1 n kb =1 m ia n kb between any atom m of active residue i in protein A (m ia ) and any atom n of both active and passive residues k (N res total) of protein B (n kb ) and inversely for protein B. For each active residue (i) in A, restraint to any active or passive residue (k) in B, over all atoms and vice versa. Dominguez et al (2003) J Am Chem Soc

38 HADDOCK calculation strategy (i) Randomization of orientations and rigid body energy minimization (ii) Semi-rigid simulated annealing (iii) Refinement with explicit solvent Models clustered by interaction energies (E elec, E vdw, E AIR ) and average buried surface area Lowest energy cluster with the highest buried surface area assumed to be correct Can also add RDCs, inter-molecular NOEs and radius of gyration term to prevent expansion at the interface (Clore(2003) JACS )

39 Sec5: chemical shift mapping data (active) and nearest neighbours (passive) Ral: 3 mutations published (active) switch regions (passive) Ral: model based on Ras Lowest energy/highest buried SA cluster Second lowest was 180 rotation Ral Sec5

40 Haddock-derived model 2.1Å Crystal Structure Ral Sec5 Ral Sec5

41 Acknowledgements Darerca Owen Louise Hopkins HR1b/Rac complex Rakhee Modha Daniel Nietlispach Ernest Laue Cdc42/PAK Angela Morreale Meenakshi Venkatesan Sec5 Jacques Camonis Medical Research Council NMR Facility: BBSRC Wellcome Trust